Nvidia and AMD Lead the AI Market. 1 Chip Stock Has Far Greater Upside.

The article favors Nvidia over AMD, arguing Nvidia's faster growth and lower relative valuation offer greater upside despite AMD's roughly 200% share gain in 2026 versus Nvidia's 30%. Nvidia's Q2 revenue rose 106% year over year to $96 billion, including data-center revenue up 117% to $89 billion; AMD revenue grew 50% to $11.5 billion, with data-center revenue up 107% to $6.7 billion. The author says AMD's AI catalysts are already reflected in its valuation, while Nvidia forecasts 70% revenue growth in 2027 and plans to launch Vera Rubin chips in the second half of this year.

Analysis

The article’s relative-value case is directionally plausible but its “three times the share price” comparison is not valuation evidence: per-share prices say nothing about ownership value or earnings multiple. Before trading the conclusion, verify current forward earnings estimates, multiples, and the revisions embedded in each stock; the article supplies none. The more useful question is whether AMD can convert announced capacity and customer relationships into shipped accelerators with attractive economics, not whether the announcements are already “priced in.” Capacity agreements are not equivalent to delivered revenue, utilization, or durable margins.

NVDA’s inference-efficiency claims could expand the economics of AI workloads and support more total demand, rather than simply crowding AMD out. Conversely, lower tokens-per-dollar may reduce compute required per task, while hyperscalers’ custom silicon and bargaining power pressure merchant-chip pricing. AMD remains a credible second source if buyers value supply diversification; that can benefit AMD even without parity in software or performance. Both companies remain exposed to the same hyperscaler capex cycle and advanced packaging/memory constraints, so the pair is not a clean hedge against an AI-spending reversal.

Near term, relative positioning and estimate revisions matter more than architecture headlines. Over 1–3 months, track shipment schedules, customer qualification, and data-center revenue/margin guidance. Over 6–18 months, watch whether inference efficiency broadens workloads and whether AMD gains repeat orders. The thesis weakens if AMD converts commitments into accelerating profitable shipments, or if NVDA guidance/revisions roll over; it strengthens if AMD’s expected earnings fail to catch up with its valuation while NVDA sustains growth.

Key Decisions for Investors

• Do not trade the article’s per-share-price argument. First compare updated forward P/E or EV/earnings against consensus EPS revisions and verify how much announced AMD business is reflected in estimates.
• For a 3–6 month relative-value expression, consider a modest beta-adjusted long NVDA / short AMD pair only if the valuation gap remains wide and estimate revisions continue to favor NVDA. Define risk by relative performance and reassess if AMD’s data-center growth and margins exceed NVDA’s trajectory; this pair still carries shared AI-capex risk.
• Treat AMD’s OpenAI capacity agreement and reported Meta-related demand as conversion watch items, not booked earnings: seek confirmation of shipment timing, purchase commitments, utilization, and segment economics before increasing exposure.
• Monitor NVDA architecture benchmarks and customer deployment economics alongside hyperscaler capex commentary. A clear AMD qualification win or NVDA guidance/revision deterioration would falsify the relative-long thesis; a capex slowdown could hurt both legs and warrants reducing gross exposure.